While only including tillage treatments with residue incorporation establishes systems with similar residue input levels, it arguably poorly reflects farmers’ predominant practices in mixed crop-livestock farming systems – especially in sub-Saharan Africa and South Asia – in which residues tend to be exported from fields for feed, fuel, housing materials or other purposes . As such, the applicability of meta-analytical results to smallholder farming conditions in either sub-Saharan Africa and South Asia may be questioned. Given the large variation in crop management practices that result from differences in the scale of farming operations, the nature of farm enterprises and cropping patterns in different farming systems, one may therefore ask: Does the presentation of average results from ‘global meta-analyses’ in agronomy make sense? Our case studies show the ways in which the practical value of meta-analyses to provide comprehensive evidence on topics of development relevance is undermined by the social construction of treatment categories that may be decoupled from the conditions faced by farmers themselves.Most meta-analyses reviewed in this study used primary data from small-plot agronomic trials. The problems associated with extrapolating results from small plot experiments to whole fields, cropping systems and farming systems have however been widely acknowledged . These problems also affect meta-analysis. Many manage multiple separate fields – each of which may be environmentally heterogeneous – across landscapes. Farmers may therefore not be able to rigorously and evenly implement recommended crop management practices across fields and farm units with the same precision as researchers managing small-plot trials. This therefore casts some doubt about the usefulness of data from small-plot trials. Kravchenko et al. ,for example,blackberries in containers demonstrated that yield results from small-plot OA experiments were not always consistent with field-scale measurements of the same treatments.
Caution is therefore needed when extrapolating results from small-plot research to the field, farming system, landscape and global levels. These problems are most apparent in the OA case study. Badgley et al. , for example, extrapolated OA yield responses from plot studies to the global agricultural system, concluding that OA could feed the world’s population with nitrogen requirements supplied in situ by legumes, without expanding the footprint of agriculture. Connor conversely pointed out that soil moisture deficits would likely constrain the productivity of legumes in arid environments. He also noted that rotations with legumes may also not be feasible where legumes are less profitable or important than other crops for income generation and food production. Assessing productivity on a yield per unit of time basis, rather than yield alone, may therefore be an appropriate alternative in such comparisons . Leifeld also referenced landscape-scale considerations when contesting data presented by Ponisio et al. . He contended that OA is unable to cope with high-fecundity and rapidly dispersing pests, which could result yield losses more severe than observed in isolated, small-plot experiments. Leifeld also evoked ‘Borlaug hypothesis’ arguments that low-yielding farming systems may require the conversion of natural ecosystems to meet expanding food demand, thereby negatively affecting biodiversity. Ponisio and Kremen countered with evidence of the positive effects of organic and ecologically managed farmland on pest suppression at the landscape scale. They also highlighted the study of Meyfroidt et al. , who showed that higher yields and profitability can also drive agricultural expansion and deforestation under conventional practices. Considering the complexity of these problems, Brandt et al.proposed that bias could be reduced and science quality increased if researchers using meta analysis make their research protocols and intended methods publically available, for example, through online posting or journal publication, prior to undertaking meta-analysis. ‘Pre-registration’ of planned studies may be a logical suggestion , though it implies serious changes in research practice and re-thinking of how journals accept papers and conduct peer-review. This proposition has therefore not yet been widely applied in agronomy or other disciplines.
While there is no easy answer to how to rectify this conundrum, our review presents and important step in challenging underlying assumptions that meta-analysis can provide definitive and unifying conclusions as proposed by Garg et al. , Borenstein et al. , Rosenthal and Schisterman and Fisher .Agricultural expansion is the main cause of tropical deforestation , highlighting the trade offs among ecosystem services such as food production, carbon storage, and biodiversity preservation inherent in land cover change . Expansion of intensive agricultural production in southern Amazonia, led by the development of specific crop varieties for tropical climates and international market demand , contributed one third of the growth in Brazil’s soybean output during 1996–2005 . The introduction of cropland agriculture in forested regions of Amazonia also changed the nature of deforestation activities; forest clearings for mechanized crop production are larger, on an average, than clearings for pasture, and the forest conversion process is often completed in o1 year . How this changing deforestation dynamic alters fire use and carbon emissions from deforestation in Amazonia is germane to studies of future land cover change , carbon accounting in tropical ecosystems , and efforts to reduce emissions from tropical deforestation . Fires for land clearing and management in Amazonia are a large anthropogenic source of carbon emissions to the atmosphere . Deforestation fires largely determine net carbon losses , because fuel loads for Amazon deforestation fires can exceed 200 Mg C ha1 . Reductions in forest biomass from selective logging before deforestation are small, averaging o10 Mg C ha 1 . In contrast, typical grass biomass for Cerrado or pasture rarely exceeds 10 Mg C ha 1 and is rapidly recovered during the subsequent wet season . Yet, the fraction of all fire activity associated with deforestation and combustion completeness of the deforestation process remain poorly quantified . Satellite fire detections have provided a general indication of spatial and temporal variation in fire activity across Amazonia for several decades . However, specific information regarding fire type or fire size can be difficult to estimate directly from active fire detections because satellites capture a snapshot of fire energy rather than a time-integrated measure of fire activity .
Overlaying active fire detections on land cover maps provides a second approach to classify fire type. Evaluating fire detections over large regions of homogenous land cover can be instructive , but geolocation errors and spurious fire detections may complicate these comparisons, especially in regions of active land cover change and high fire activity such as Amazonia . Finally, postfire detection of burn-scarred vegetation is the most data-intensive method to quantify carbon emissions from fires. Two recent approaches to map burn scars with Moderate Resolution Imaging Spectroradiometer data show great promise for identifying large-scale fires , yet neither algorithm is capable of identifying multiple burning events in the same ground location typical of deforestation activity in Amazonia. Deriving patterns of fire type, duration and intensity of fire use, and combustion completeness directly from satellite fire detections provides an effi- cient alternative to more data and labor-intensive methods to estimate carbon emissions from land cover change. We assess the contribution of deforestation to fire activity in Amazonia based on the intensity of fire use during the forest conversion process,blackberry container measured as the local frequency of MODIS active fire detections. High confidence fire detections on 2 or more days in the same dry season are possible in areas of active deforestation, where trunks, branches, and other woody fuels can be piled and burned many times. Low-frequency fire detections are typical of fires in Cerrado woodland savannas and for agricultural maintenance, because grass and crop residues are fully consumed by a single fire. The frequency of fires at the same location, or fire persistence, has been used previously to assess Amazon forest fire severity , adjust burned area estimates in tropical forest ecosystems , and scale combustion completeness estimates in a coarse-resolution fire emission model . We build on these approaches to characterize fire activity at multiple scales. First, we compare the frequency of satellite fire detections over recently deforested areas with that over other land cover types. We then assess regional trends in the contribution of high frequency fires typical of deforestation activity to the total satellite-based fire detections for Amazonia during 2003–2007. Finally, we compare temporal patterns of fire usage among individual deforested areas with different post clearing land uses, based on a recent work to separate pasture and cropland following forest conversion in the Brazilian state of Mato Grosso with vegetation phenology data . The goals of this research are to test whether fire frequency distinguishes between deforestation fires and other fire types and characterize fire frequency as a function of post clearing land use to enable direct interpretation of MODIS active fire data for relevant information on carbon emissions.We analyzed active fire detections from the MODIS sensors aboard the Terra and Aqua satellite platforms to determine spatial and temporal patterns in satellite fire detections from deforestation in Amazonia during this period.
Combined, the MODIS sensors provide two daytime and two night-time observations of fire activity. Figure 1 shows the location of the study area and administrative boundaries of the nine countries that contain portions of the Amazon Basin. For data from 2002–2006, the date and center location of each MODIS active fire detection, satellite , time of overpass, 4 micron brightness temperature , and confidence score were extracted from the Collection 4 MODIS Thermal Anomalies/Fire 5-min swath product at 1-km spatial resolution . Beginning in 2007, MODIS products were transitioned to Collection 5 algorithms. Data for January 1–November 1, 2007 were provided by the Fire Information for Resource Management System at the University of Maryland, College Park based on the Collection 5 processing code. Seasonal differences in fire activity north and south of the equator related to precipitation were captured using different annual calculations. North of the equator, the fire year was July–June; south of the equator, the fire year was January–December. Our analysis considered a high-confidence subset of all MODIS fire detections to reduce the influence of false fire detections over small forest clearings in Amazonia . For daytime fires, only those 1-km fire pixels having 4330 K brightness temperature in the 4-mm channel were considered. This threshold is set based on a recent work to identify true and false MODIS fire detections with coincident high-resolution satellite imagery , comparisons with field data , and evidence of unrealistic MODIS fire detections over small historic forest clearings in Mato Grosso state with 420 days of fire detections per year in 3 or more consecutive years, none of which exceeded 330 K during the day. Daytime fire detections 4330 K correspond toa MOD14/MYD14 product confidence score of approximately 80/100. The subset of high-confidence fires includes all night-time fire detections, regardless of brightness temperature. Differential surface heating between forested and cleared areas during daylight hours that may contribute to false detections should dissipate by the 22:30 or 01:30 hours local time overpasses for Terra and Aqua, respectively. Subsequent references to MODIS fire detections refer only to the high-confidence subset of all 1-km fire pixels described earlier.The simple method we propose for separating deforestation and agricultural maintenance fires is based on evidence for repeated burning at the same ground locations. The spatial resolution of our analysis is de- fined by the orbital and sensor specifications of the MODIS sensors and the 1-km resolution bands used for fire detection. The geolocation of MODIS products is highly accurate, and surface location errors are generally o70 m . However, due to the orbital characteristics of the Terra and Aqua satellite platforms, the ground locations of each 1-km pixel are not fixed. We analyzed three static fire sources from gas , mining , and steel production in South America to identify the spatial envelope for MODIS active fire detections referencing the same ground location. Over 98% of the high-confidence 2004 MODIS active fire detections from Terra and Aqua for these static sources were within 1 km of the ground location of these facilities. Therefore, we used this empirically derived search radius to identify repeated burning of forest vegetation during the conversion process. High-frequency fire activity was defined as fire detections on two or more days within a 1-km radius during the same fire year.